ABSTRACT It is easy for children to overeat in today's food environment, leading to weight gain. However, not all children are obese, suggesting that there is individual variability in how people perceive and interact with the current food environment. Research on obesity shows that overweight and obese people, and those that gain weight, compared to healthy weight people, have greater reward sensitivity, as demonstrated by increased BOLD response in neural areas associated with reward, evaluation and taste in response to the taste of food. Since relationships are learned through associative conditioning between environmental cues and eating, it is possible that this individual variability in how people interact with the food environment could be explained by different rates of learning between food cues in the environment and tastes of food. Thus, an alternative hypothesis, is that people that gain weight could be more likely to learn the relationships between cues in the environment predicting food (cue-reward learning). In this study, we propose to evaluate whether reward sensitivity, cue-reward learning, or both are differentially associated with children at risk for obesity and those not at risk for obesity, and to evaluate whether these neural mechanisms predict weight gain over time. We propose to recruit and scan 66 healthy weight 8-10 year old children at high-risk for obesity (HR: two overweight/obese parents) and 66 healthy weight age matched children at low-risk for obesity (LR; two lean parents) using functional MRI evaluating BOLD response to a cue-reward learning paradigm which pairs innocuous cues with chocolate milkshake and tasteless saliva. In particular, we are interested in evaluating whether reward sensitivity and cue-reward learning in the amygdala, insula, hippocampus and striatum will differentiate the HR from the LR children, and whether these two mechanisms predict weight gain and eating over two years. We will achieve this goal through the following three primary aims: 1) Compare the rate of cue-reward learning to cues paired with chocolate milkshake compared to cues paired with tasteless saliva between HR and LR children, 2) Compare the reward sensitivity, as measured by peak BOLD response, between HR and LR children 3) To evaluate cue-reward learning and reward sensitivity to hedonic taste as predictors of weight gain and eating over time in the children (HR and LR). This program of research tests a novel hypothesis regarding overeating and the development of obesity in children, and could provide critical data on individual vulnerabilities to overeating for further research. Furthermore, this study could provide mechanisms for intervention with regards to cue-reward learning in children, to ultimately prevent obesity in youth.